This study will test the hypothesis that subjects with HIV infection, who are also taking nucleoside analogues, have an abnormal anaerobic threshold to exercise as demonstrated by lactate generation. The comparisons will be to normal age matched volunteers, and if possible, with HIV positive subjects not taking nucleoside analogues. This study will try to establish in vivo that nucleoside analogues, even in asymptomatic HIV positive subjects, alter the mitochondrial function of muscle.